摘要
传统的网络新闻信息传播流行度预测系统存在预测准确率低、预测响应时间较长等问题,为此提出了基于LSTM算法的网络新闻信息传播流行度预测系统。分析信息传播对网络拓扑结构产生的影响,加入用户对新闻信息的兴趣程度和信息价值等相关因素,获取用户节点状态转换函数。将粒子群算法中的适应度函数设定为求解依据,修改LSTM算法中的阈值和权值,建立网络新闻信息传播流行度预测模型。在此基础上,进行系统设计,系统主要是由网络新闻抓取和存储模块、网络新闻展示等模块组成,分别对各个模块的功能进行了详细地分析和介绍。仿真实验结果表明,所设计系统不仅能够有效提升网络新闻信息传播流行度预测准确率,同时还能够降低预测响应时间以及预测费用。
The traditional network news information dissemination popularity prediction system has the problems of low prediction accuracy and long prediction response time.Therefore,a network news information dissemination popularity prediction system based on LSTM algorithm is proposed.The influence of information transmission on network topology is analyzed,and relevant factors such as users’interest in news information and information value are added to obtain the user node state transition function.The fitness function in the particle swarm optimization algorithm is set as the basis for solving the problem,and the threshold and weight in the LSTM algorithm are modified to establish a prediction model for the popularity of network news and information communication.On this basis,the paper carries on the system design.The system is mainly composed of network news capture and storage module,network news display module and so on,respectively.The paper carries on the detailed analysis and introduction to each module’s function.The simulation results show that the designed system can not only effectively improve the prediction accuracy of the popularity of network news and information transmission,but also reduce the prediction response time and prediction cost.
作者
门玉霞
MEN Yuxia(Zigong First Municipal People’s Hospital,Zigong 643000,China)
出处
《微型电脑应用》
2022年第11期152-155,共4页
Microcomputer Applications
